10.

626 Electrochemical Energy Systems

Spring 2014 MIT, M. Z. Bazant

Problem Set 5 – Transport

1. Diffusion in Linear Sweep Voltammetry. The transient response of a cell is limited by

the diffusion of a reactant,
∂c ∂2c
=D 2
∂t ∂x
in a semi-infinite domain x > 0. Relative to open circuit conditions where c(x) = c0 , the
change in cell voltage is small enough to linearize the Nernst equation,
kT Δc
V − V0 =
ne c0
during a linear voltage sweep, V (t) = V0 + St, where S is the sweep rate. The current is given
by
∂c
I(t) = −neAD (x = 0, t)
∂x
∂c
(a) Show that u(x, t) = ∂t satisfies the diffusion equation, and solve by inspection.
(b) Derive c(x, t).
(c) Derive and plot the current-voltage relation, I(V, S).

2. Membraneless fuel cell. Two streams of liquid electrolyte containing initially separated
reactants flow through a porous separator with uniform velocity u between parallel plate
electrodes of length L separated by 2H. The inlets each cover half of the channel width, H.
Fuel A enters in the stream near the anode, and oxidant B and product AB enter in the other
stream near the cathode.

The electrochemical reactions are:

anode: A → A+ + e− , ΔφΘ = ΔφΘ

a
cathode: B + A+ + e− → AB, ΔφΘ = ΔφΘ
c
net reaction: A + B → AB

where standard electrode potentials ΔφΘ are given for the Faradaic reactions (with all re­
actants at 1 M concentration). Assume that the incoming streams contain enough A+ ions

1
 that their concentration remains roughly constant throughout the cell. Let c̄A , c̄B , and c̄AB
be the concentrations at the inlets, and DA , DB , and DAB the diffusivities. Assume c̄A « c̄B
and fast reactions, so that the limiting current Ilim is controlled only by the transport of fuel
A to the anode. The cell is operated in steady state.

(a) Determine the fuel utilization of species A, γ(I, u), (fraction of incoming fuel consumed
by reactions) at a given current I (per width into the page).
(b) Make a scaling argument to estimate the minimum fluid velocity umin required to avoid
fuel crossover (i.e. A reaching the cathode or B reaching the anode before exiting the
cell, thereby reducing the voltage).
(c) Assuming u » umin and dilute solutions, determine the open circuit voltage, V0 .
(d) For u » umin , neglect axial diffusion and derive Ilim (u) by setting cA = 0 on the anode.
(e) As a first approximation for I < Ilim , assume a uniform concentration cA of fuel A over
the anode surface, given by
cA I
=1−
c̄A Ilim
and calculate the power P (I, u) of the fuel cell.
(f) For the typical case, Ṽ0 = eV0 /kB T » 1, estimate the maximum power Pmax (u) of the
fuel cell at a given flow rate, and show that Pmax γ ≈ constant.

3. Electrocatalytic impedance. The interfacial impedance of electrocatalytic reactions

involves not only charge transfer, but also surface adsorption and surface diffusion of reactants
to the triple phase boundary (TPB,) where the electrolyte, electrode, and reactant phases
meet. In a solid oxide fuel cell (Figure below), oxygen gas adsorbs and dissociates on the
LSM electrode surface and diffuses to the YSZ electrolyte TPB where charge transfer occurs.
At high frequency, charge transfer impedance dominates, and the oxygen surface coverage
remains unperturbed, as in Homework #3. Here we consider lower frequencies, where charge
transfer is fast, and the surface adsorption/diffusion process is rate limiting.

(a) Set up the problem.

i. Assume a Langmuir adsorption model. 1 Let c̃ be the dimensionless surface cov­
erage, scaled to the maximum surface concentration cs (sites/area). Let p be the
partial pressure (in atm) of oxygen gas. Let Ka be the equilibrium constant of the
adsorption reaction, O2 (gas) → 2 O (surface), and kd be the backward (desorption)
rate constant. What is the net adsorption rate Ra (number/time) per surface site?
What is the equilibrium coverage c̃eq where Ra = 0?
ii. Let Δc̃(x, t) = c̃(x, t) − c̃eq be the surface coverage perturbation in response to an
applied voltage ΔV (t). Consider one-dimensional diffusion in a coordinate x away
from the TPB (at x = 0). Linearize for |Δc̃| « c̃eq , and derive the linear response
equation
∂Δc̃ ∂ 2 Δc̃
= Ds − kΔc̃
∂t ∂x2
where Ds is the surface diffusivity of adsorbed oxygen. What is k?
iii. Let A be the electrode area and lT P B the TPB density (length/area). Relate ΔV (t)
and ΔI(t) (change in total current) to Δc̃(0, t) and ∂Δ˜ c
∂x (0, t). Consider AC forcing
at frequency ω, and relate the impedance Z(ω) to the complex amplitude Δc̃∗ (x).
1
A lattice gas or ideal solution of particles and vacancies on the surface with two excluded sites in the transition
state.

2
 (b) Consider a semi-infinite catalytic surface.
i. Derive Z = Z∞ (ω). 2
ii. Sketch the Nyquist plot. [Hint: square root of a semi-circle.]
iii. Show and explain why the high frequency limit ω » k is a Warburg element.
iv. Show and explain why the low frequency limit ω « k is a resistance for quasi-steady
diffusion across a boundary-layer thickness (Fig. 1).
(c) Now consider a finite catalytic surface of width 2L with symmetry about x = L, i.e.
∂Δc̃ 2
∂x (L, t) = 0. Let k̃ = kL /Ds be the Thiele modulus.
i. Derive Z(ω).
ii. For k̃ » 1, show and explain why Z(ω) ∼ Z∞ (ω).
iii. For k̃ « 1, show and explain why Z(ω) reduces to a bounded Warburg element for
ω » k.
iv. For k̃ « 1, show and explain why Z(ω) ∝ Z∞ (ω) for ω « k with a much larger
low-frequency resistance (i.e. poor catalytic activity).

Courtesy of Yeqing Fu. Used with permission.

Figure 1: Heterogeneous electrocatalysis of oxygen reduction in a solid oxide fuel cell. (Courtesy
of Yeqing Fu)

2
This form was introduced by Gerischer 1951 and carries his name.

3
MIT OpenCourseWare
http://ocw.mit.edu

10.626 Electrochemical Energy Systems

Spring 2014

For information about citing these materials or our Terms of Use, visit: http://ocw.mit.edu/terms.